Prediction of complex flow patterns in intracranial atherosclerotic disease using computational fluid dynamics.

Although carotid and vertebral intracranial atherosclerotic disease (ICAD) can lead to both hemodynamic insufficiency and thromboembolism, its fluid dynamic properties remain undefined because of its intricate features and complex three-dimensional geometry. We used computational fluid dynamic (CFD) analysis to model the hemodynamics of symptomatic ICAD lesions. Nine ICAD lesions (six carotid, two vertebral, one middle cerebral) underwent high-resolution catheter-based digital rotational angiography. The reconstructed three-dimensional volumes of the target lesions were segmented and used to generate hybrid computational meshes. Dynamic pulsatile CFD analysis was performed using a non-Newtonian shear-dependent model of blood's viscosity. CFD results revealed complex flow patterns within ICAD lesions with midstenotic shear rates of greater than 19,000/s, sufficiently high to induce high-shear platelet activation. Vorticity and helicity within the stenoses were followed by sudden deceleration with formation of vortex cores. Pressure gradients were significant mostly at greater than 75% stenosis with a mean time-averaged drop of 27.2 +/-17.8 mmHg. Unlike the smoothly-varying helicity imparted by the three-dimensional anatomy of the intracranial circulation, poststenotic regions of ICAD lesions showed significant and rapidly fluctuating helicity and vorticity patterns, which may contribute to the propagation of platelets activated by the high shear region within the stenosis throat. Stent angioplasty restored the hemodynamic profile of ICAD lesions to within contralateral controls. Patient-based symptomatic ICAD lesions studied using CFD analysis appear to harbor a hemodynamically pathological environment that favors the activation, aggregation and distal embolization of platelets and is reversed by endovascular stent angioplasty.